Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Product of two integrals

  1. Dec 15, 2007 #1
    Product of two integrals....

    In proving a theorem, my DE textbook uses an unfamiliar approach by stating that
    the product of two integrals = double integral sign - the product of two functions - dx dy

    i hope my statement is descriptive enough. My question is, what's the proof to this?
  2. jcsd
  3. Dec 15, 2007 #2


    User Avatar
    Science Advisor
    Homework Helper

    You mean, why is it true that
    [tex]\left( \int f(x) dx \right) \left( \int g(x) dx \right) = \iint f(x) g(y) dx dy?[/tex]
    Because that's simply by the property that we can put any "constant" (that is, anything which does not depend on the integration variable) outside or inside an integral as we like. We can obviously rewrite
    [tex]\left( \int f(x) dx \right) \left( \int g(x) dx \right) = \left( \int f(x) dx \right) \left( \int g(y) dy \right)[/tex]
    by renaming a dummy variable in the second integral. Now you see that the first part does not depend on y at all (just on x, but that's a different variable!), so we can take the entire first integral inside the second integration,
    [tex]\left( \int f(x) dx \right) \left( \int g(x) dx \right) = \int \left( \int f(x) dx \right) g(y) dy[/tex]
    and since g(y) doesn't depend on x we can take that inside the x-integration (and then of course drop the brackets, as they don't clarify anything anymore).

    Hope that answers the question.
  4. Dec 15, 2007 #3

    Gib Z

    User Avatar
    Homework Helper

    Rewriting the dummy variable only yields the same result in Definite integration. I'm sure CompuChip meant this, writing bounds on all those integral signs wouldn't have been fun.
  5. Dec 16, 2007 #4


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    It is a property of Fubini's theorem that assuming the double integral over the absolute value of f converges, then it will yield the same value as the "nested" integral computations, irrespective of nesting order.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?